Modal Seismic Control of Building Frames Using ANN

Abstract eng:
Target reduction of seismic response of a multi-storey frame with a neurocontroller (using Artificial Neural Network) is presented. The seismic response of the frame is controlled by controlling the significant modal contributions (modal control) to the overall response. ANN with a feed-forward architecture is used to construct, train and test the performance of the neurocontroller. Control methodology uses two sets of neural nets wherein output of first set acts as part of the input to the second set. Number of neural nets to be trained depends upon the number of modal response being controlled. The neurocontroller is designed to provide a target reduction of response by taking into account the time delay effect also. Inputs to this scheme are the measured accelerations only at few selected points of the structure, and the ground acceleration. The neural nets are trained for the synthetically generated input-output data with the help of simulated earthquake records having different frequency compositions. The effectiveness of the control scheme is tested for both known and unknown (El Centro and Treasure Island earthquake records) problems for a ten storey building frame. Results of the study show that the control scheme is highly effective in controlling both displacement and acceleration responses of the frame for the unknown El Centro and Treasure Island earthquake excitations.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket